Flexibility-Oriented Collaborative Planning Model for Distribution Network and EV Parking Lots Considering Uncertain Behaviour of EVs

Abstract

Increasing grid integration of intermittent renewable energy sources (RESs) and plug-in electric vehicles (PEVs) with uncertain behaviours have necessitated enhancing the flexibility requirements of distribution networks. Thus, in the state-of-the-art distribution network expansion planning (DNEP) models, both flexibility requirements and high penetration of RESs and PEVs should be taken into consideration. In this respect, a novel collaborative planning model for power distribution network (PDN) and plug-in Electric Vehicle Parking Lots (EVPLs) is proposed in this paper, which leverages sizing, siting, and operation of EVPLs to enhance the distribution network flexibility. Also, to model the uncertain traffic flow of PEVs, a new model is proposed and is utilized to obtain a preliminary dispatch of PEV charging and, in turn, an estimated EVPL demand. Afterwards, this estimated demand is fed into the collaborative planning model to obtain the optimal expansion planning solution for PDN, and the size and location of EVPLs. Nonetheless, to provide the network operator with more flexibility sources, it is assumed that the operator can reschedule the charging pattern of some PEVs by compensating the EVPL owners for the difference in retail electricity prices of various hours. Finally, to illustrate the effectiveness of the proposed model, it is implemented on a test network, and the obtained results are discussed.

Planning of electric vehicle charging stations and distribution system with highly renewable penetrations

“…However, in cases where the driving patterns are different, probabilistic models might be more suitable. Furthermore, [1,2,[4][5][6][7] utilised commercial solvers to obtain the final solutions. However, it has been shown in various applications that genetic algorithms can achieve comparable results in similar or less time [8][9][10].…”

“…Additionally, for the randomness of EV driving behaviours and wind generation, [14] presented a scenario-based stochastic programme to model the uncertain nature of EV charging demand. In [5,11,13,14] the EV charging model is based on Monte-Carlo simulations (MCSs) of multidimensional joint probability distribution, and the representative scenarios with high-precision results can only be obtained when the size of the initial generated scenarios is large enough. However, the large initial scenario set will considerably increase the computational cost of the scenario reduction procedure, thus decreasing the efficiency of stochastic optimisation planning.…”

“…In contrast to the isolated distribution systems in [1][2][3][4][5]11,15], a coupled system is considered here, which consists of a traffic network and a power distribution network. Unlike [4,12,13,15], the power distribution network herein takes high wind power penetration into account.…”